RE: [Xen-users] KVM Tests

["Petersson, Mats" <Mats.Petersson@xxxxxxx>]

> -----Original Message-----
> From: xen-users-bounces@xxxxxxxxxxxxxxxxxxx 
> [mailto:xen-users-bounces@xxxxxxxxxxxxxxxxxxx] On Behalf Of 
> Artur Baruchi
> Sent: 02 March 2007 18:30
> To: Xen-users@xxxxxxxxxxxxxxxxxxx
> Subject: [Xen-users] KVM Tests
> 
> Hi Guys,
> 
> Somebody knows if KVM improve the performance of Xen?! 
> Anybody have tested it?

KVM and Xen are competing and non-compatible in a direct usage way. You
may be able to run Xen as a virtual machine on top of KVM (possibly),
but I doubt that you'll be able to do anything particularly useful by
such a combination. 

KVM may be a bit more efficient in some respects, but most of the
overhead in any (full) virtualization method is the device modeling, and
there's no great difference in these two between KVM and Xen. 

Xen's para-virtual model should beat any full virtualization technology
available today - future will tell if the nested page-table support will
make full virtualization as fast than the PV model with software-based
page-table support. Page-table management is another big hit point in
all virtualization, but I doubt that KVM has any great advantage over
Xen in this respect.

--
Mats
> 
> Thanks in advance,
> 
> Artur Baruchi
> 
> _______________________________________________
> Xen-users mailing list
> Xen-users@xxxxxxxxxxxxxxxxxxx
> http://lists.xensource.com/xen-users
> 
> 
> 



_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users